A multi-dimensional approach to predicting speech quality using a physiologically motivated model of the Cochlea

Author

Sen, D. ; Lu, Wenliang

Author_Institution

Sch. of Electr. Eng. & Telecommun., Univ. of New South Wales, Sydney, NSW, Australia

fYear

2011

fDate

22-27 May 2011

Firstpage

5144

Lastpage

5147

Abstract

This paper introduces a framework to measure speech quality in a multi-dimensional space. A hydro-mechanical Cochlear model is used to convert speech into the perceptual domain, where salient features are extracted to detect temporal and frequency localized distortions. Overall quality evaluation is subsequently achieved using a weighted sum of the detectability of individual distortions.

Keywords

ear; feature extraction; speech processing; cochlea physiologically motivated model; frequency localized distortion; hydromechanical cochlear model; multidimensional approach; perceptual domain; quality evaluation; salient feature extraction; speech conversion; speech quality measurement; speech quality prediction; temporal detection; Correlation; Distortion measurement; Feature extraction; Noise; Nonlinear distortion; Psychoacoustic models; Speech; Diagnostic Acceptability Measure; Objective measurement of speech quality;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on

Conference_Location

Prague

ISSN

1520-6149

Print_ISBN

978-1-4577-0538-0

Electronic_ISBN

1520-6149

Type

conf

DOI

10.1109/ICASSP.2011.5947515

Filename

5947515